国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

by which these objects were fixed to the soil of the tomb, since it seems likely
that burial objects were placed with the dead in a certain arrangement. Pairs
of animals on a square plaque are also found at the Chinese frontier, and were
probably meant for the same purpose. Finally, the subject chosen for this type
of figure seems to prove our supposition, since a soliped, an animal that was
the food of the nomads procured either by hunting or by rearing, was nearly
always reproduced.

The roe is executed in many different ways. This animal, either frightened or
resting, is from an artistic point of view, the most successfully treated because
of its naturalism. Startled, it is about to rise, or is alert with head turned back,
or with head thrust forward scents an enemy. The conventional arrangement
of the legs, drawn together or one above the other, has already been found
in the oldest art of the Scythians on the shores of the Black Sea.

Many types of resting roe are found in pairs. This is true of the very attractive
figure with back-turned head (Plate XXXIII no. 2). The Minussinsk perfora-
tions are entirely lacking, the holes which can be seen in the reproduction being
due to the decay of the surface. The body is only slightly articulated, a type
that does not seem possible before the end of the I millennium. The same date
applies to the roe with head straight froward (Pl. XXXIII, no. 3). This figure
seems to express a certain calm repose, rather than an animal on the alert and
listening. There is no hollow here between the legs and belly, but instead there
are two holes, meant perhaps for a transversal peg.

The standing animal is unusual (Plate XXXIII no. 4), but even here we find
a hollow so that the legs have only an outside surface.

With Plate XXXIII no. 5 we are approaching mechanically repeated forms,
that is, we are entering the II millennium. The tongue can be seen in this example,
a peculiarity only to be looked for when a dying or dead animal is figured. But
here the straining neck of the roe gives it the appearance of life. Plate XXXIII
no. 6 shows us once again the characteristics of a period after 1000. The careless
casting of the two halves, the edges of which have not been cleaned, the unmodeled
surface, so simplified that it is primitive in form, all point to a late date. The
tongue is visible again in a variation of the last example (Plate XXXIV no. 1).
It is impossible to determine whether the animal is about to spring forward or
is collapsing on the ground. We find a geometric ornamentation on a small
animal that originally was reproduced in a naturalistic style (Plate XXXIV
no. 2). The decoration is composed of round spiral ribbons on the joints, and
an angular spiral on the belly. So conventional a rendering seems only possible
in the II millennium.

Plate XXXIV no. 3 is such a combination of incongruous animal forms that
it entirely lacks naturalism. Only the legs seem to be those of a roe. The
elongated body, the upturned tail, the small ears, all belong to a wild animal.